2 * UEFI Common Platform Error Record (CPER) support
4 * Copyright (C) 2010, Intel Corp.
5 * Author: Huang Ying <ying.huang@intel.com>
7 * CPER is the format used to describe platform hardware error by
8 * various tables, such as ERST, BERT and HEST etc.
10 * For more information about CPER, please refer to Appendix N of UEFI
11 * Specification version 2.4.
13 * This program is free software; you can redistribute it and/or
14 * modify it under the terms of the GNU General Public License version
15 * 2 as published by the Free Software Foundation.
17 * This program is distributed in the hope that it will be useful,
18 * but WITHOUT ANY WARRANTY; without even the implied warranty of
19 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
20 * GNU General Public License for more details.
22 * You should have received a copy of the GNU General Public License
23 * along with this program; if not, write to the Free Software
24 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
27 #include <linux/kernel.h>
28 #include <linux/module.h>
29 #include <linux/time.h>
30 #include <linux/cper.h>
31 #include <linux/dmi.h>
32 #include <linux/acpi.h>
33 #include <linux/pci.h>
34 #include <linux/aer.h>
38 static char rcd_decode_str[CPER_REC_LEN];
41 * CPER record ID need to be unique even after reboot, because record
42 * ID is used as index for ERST storage, while CPER records from
43 * multiple boot may co-exist in ERST.
45 u64 cper_next_record_id(void)
47 static atomic64_t seq;
49 if (!atomic64_read(&seq))
50 atomic64_set(&seq, ((u64)get_seconds()) << 32);
52 return atomic64_inc_return(&seq);
54 EXPORT_SYMBOL_GPL(cper_next_record_id);
56 static const char * const severity_strs[] = {
63 const char *cper_severity_str(unsigned int severity)
65 return severity < ARRAY_SIZE(severity_strs) ?
66 severity_strs[severity] : "unknown";
68 EXPORT_SYMBOL_GPL(cper_severity_str);
71 * cper_print_bits - print strings for set bits
72 * @pfx: prefix for each line, including log level and prefix string
74 * @strs: string array, indexed by bit position
75 * @strs_size: size of the string array: @strs
77 * For each set bit in @bits, print the corresponding string in @strs.
78 * If the output length is longer than 80, multiple line will be
79 * printed, with @pfx is printed at the beginning of each line.
81 void cper_print_bits(const char *pfx, unsigned int bits,
82 const char * const strs[], unsigned int strs_size)
88 for (i = 0; i < strs_size; i++) {
89 if (!(bits & (1U << i)))
94 if (len && len + strlen(str) + 2 > 80) {
99 len = snprintf(buf, sizeof(buf), "%s%s", pfx, str);
101 len += snprintf(buf+len, sizeof(buf)-len, ", %s", str);
107 static const char * const proc_type_strs[] = {
112 static const char * const proc_isa_strs[] = {
118 static const char * const proc_error_type_strs[] = {
122 "micro-architectural error",
125 static const char * const proc_op_strs[] = {
126 "unknown or generic",
129 "instruction execution",
132 static const char * const proc_flag_strs[] = {
139 static void cper_print_proc_generic(const char *pfx,
140 const struct cper_sec_proc_generic *proc)
142 if (proc->validation_bits & CPER_PROC_VALID_TYPE)
143 printk("%s""processor_type: %d, %s\n", pfx, proc->proc_type,
144 proc->proc_type < ARRAY_SIZE(proc_type_strs) ?
145 proc_type_strs[proc->proc_type] : "unknown");
146 if (proc->validation_bits & CPER_PROC_VALID_ISA)
147 printk("%s""processor_isa: %d, %s\n", pfx, proc->proc_isa,
148 proc->proc_isa < ARRAY_SIZE(proc_isa_strs) ?
149 proc_isa_strs[proc->proc_isa] : "unknown");
150 if (proc->validation_bits & CPER_PROC_VALID_ERROR_TYPE) {
151 printk("%s""error_type: 0x%02x\n", pfx, proc->proc_error_type);
152 cper_print_bits(pfx, proc->proc_error_type,
153 proc_error_type_strs,
154 ARRAY_SIZE(proc_error_type_strs));
156 if (proc->validation_bits & CPER_PROC_VALID_OPERATION)
157 printk("%s""operation: %d, %s\n", pfx, proc->operation,
158 proc->operation < ARRAY_SIZE(proc_op_strs) ?
159 proc_op_strs[proc->operation] : "unknown");
160 if (proc->validation_bits & CPER_PROC_VALID_FLAGS) {
161 printk("%s""flags: 0x%02x\n", pfx, proc->flags);
162 cper_print_bits(pfx, proc->flags, proc_flag_strs,
163 ARRAY_SIZE(proc_flag_strs));
165 if (proc->validation_bits & CPER_PROC_VALID_LEVEL)
166 printk("%s""level: %d\n", pfx, proc->level);
167 if (proc->validation_bits & CPER_PROC_VALID_VERSION)
168 printk("%s""version_info: 0x%016llx\n", pfx, proc->cpu_version);
169 if (proc->validation_bits & CPER_PROC_VALID_ID)
170 printk("%s""processor_id: 0x%016llx\n", pfx, proc->proc_id);
171 if (proc->validation_bits & CPER_PROC_VALID_TARGET_ADDRESS)
172 printk("%s""target_address: 0x%016llx\n",
173 pfx, proc->target_addr);
174 if (proc->validation_bits & CPER_PROC_VALID_REQUESTOR_ID)
175 printk("%s""requestor_id: 0x%016llx\n",
176 pfx, proc->requestor_id);
177 if (proc->validation_bits & CPER_PROC_VALID_RESPONDER_ID)
178 printk("%s""responder_id: 0x%016llx\n",
179 pfx, proc->responder_id);
180 if (proc->validation_bits & CPER_PROC_VALID_IP)
181 printk("%s""IP: 0x%016llx\n", pfx, proc->ip);
184 static const char * const mem_err_type_strs[] = {
189 "single-symbol chipkill ECC",
190 "multi-symbol chipkill ECC",
198 "scrub corrected error",
199 "scrub uncorrected error",
200 "physical memory map-out event",
203 const char *cper_mem_err_type_str(unsigned int etype)
205 return etype < ARRAY_SIZE(mem_err_type_strs) ?
206 mem_err_type_strs[etype] : "unknown";
208 EXPORT_SYMBOL_GPL(cper_mem_err_type_str);
210 static int cper_mem_err_location(const struct cper_sec_mem_err *mem, char *msg)
218 len = CPER_REC_LEN - 1;
219 if (mem->validation_bits & CPER_MEM_VALID_NODE)
220 n += scnprintf(msg + n, len - n, "node: %d ", mem->node);
221 if (mem->validation_bits & CPER_MEM_VALID_CARD)
222 n += scnprintf(msg + n, len - n, "card: %d ", mem->card);
223 if (mem->validation_bits & CPER_MEM_VALID_MODULE)
224 n += scnprintf(msg + n, len - n, "module: %d ", mem->module);
225 if (mem->validation_bits & CPER_MEM_VALID_RANK_NUMBER)
226 n += scnprintf(msg + n, len - n, "rank: %d ", mem->rank);
227 if (mem->validation_bits & CPER_MEM_VALID_BANK)
228 n += scnprintf(msg + n, len - n, "bank: %d ", mem->bank);
229 if (mem->validation_bits & CPER_MEM_VALID_DEVICE)
230 n += scnprintf(msg + n, len - n, "device: %d ", mem->device);
231 if (mem->validation_bits & CPER_MEM_VALID_ROW)
232 n += scnprintf(msg + n, len - n, "row: %d ", mem->row);
233 if (mem->validation_bits & CPER_MEM_VALID_COLUMN)
234 n += scnprintf(msg + n, len - n, "column: %d ", mem->column);
235 if (mem->validation_bits & CPER_MEM_VALID_BIT_POSITION)
236 n += scnprintf(msg + n, len - n, "bit_position: %d ",
238 if (mem->validation_bits & CPER_MEM_VALID_REQUESTOR_ID)
239 n += scnprintf(msg + n, len - n, "requestor_id: 0x%016llx ",
241 if (mem->validation_bits & CPER_MEM_VALID_RESPONDER_ID)
242 n += scnprintf(msg + n, len - n, "responder_id: 0x%016llx ",
244 if (mem->validation_bits & CPER_MEM_VALID_TARGET_ID)
245 scnprintf(msg + n, len - n, "target_id: 0x%016llx ",
252 static int cper_dimm_err_location(const struct cper_sec_mem_err *mem, char *msg)
255 const char *bank = NULL, *device = NULL;
257 if (!msg || !(mem->validation_bits & CPER_MEM_VALID_MODULE_HANDLE))
261 len = CPER_REC_LEN - 1;
262 dmi_memdev_name(mem->mem_dev_handle, &bank, &device);
264 n = snprintf(msg, len, "DIMM location: %s %s ", bank, device);
266 n = snprintf(msg, len,
267 "DIMM location: not present. DMI handle: 0x%.4x ",
268 mem->mem_dev_handle);
274 static void cper_print_mem(const char *pfx, const struct cper_sec_mem_err *mem)
276 if (mem->validation_bits & CPER_MEM_VALID_ERROR_STATUS)
277 printk("%s""error_status: 0x%016llx\n", pfx, mem->error_status);
278 if (mem->validation_bits & CPER_MEM_VALID_PA)
279 printk("%s""physical_address: 0x%016llx\n",
280 pfx, mem->physical_addr);
281 if (mem->validation_bits & CPER_MEM_VALID_PA_MASK)
282 printk("%s""physical_address_mask: 0x%016llx\n",
283 pfx, mem->physical_addr_mask);
284 if (cper_mem_err_location(mem, rcd_decode_str))
285 printk("%s%s\n", pfx, rcd_decode_str);
286 if (mem->validation_bits & CPER_MEM_VALID_ERROR_TYPE) {
287 u8 etype = mem->error_type;
288 printk("%s""error_type: %d, %s\n", pfx, etype,
289 cper_mem_err_type_str(etype));
291 if (cper_dimm_err_location(mem, rcd_decode_str))
292 printk("%s%s\n", pfx, rcd_decode_str);
295 static const char * const pcie_port_type_strs[] = {
297 "legacy PCI end point",
301 "upstream switch port",
302 "downstream switch port",
303 "PCIe to PCI/PCI-X bridge",
304 "PCI/PCI-X to PCIe bridge",
305 "root complex integrated endpoint device",
306 "root complex event collector",
309 static void cper_print_pcie(const char *pfx, const struct cper_sec_pcie *pcie,
310 const struct acpi_generic_data *gdata)
312 if (pcie->validation_bits & CPER_PCIE_VALID_PORT_TYPE)
313 printk("%s""port_type: %d, %s\n", pfx, pcie->port_type,
314 pcie->port_type < ARRAY_SIZE(pcie_port_type_strs) ?
315 pcie_port_type_strs[pcie->port_type] : "unknown");
316 if (pcie->validation_bits & CPER_PCIE_VALID_VERSION)
317 printk("%s""version: %d.%d\n", pfx,
318 pcie->version.major, pcie->version.minor);
319 if (pcie->validation_bits & CPER_PCIE_VALID_COMMAND_STATUS)
320 printk("%s""command: 0x%04x, status: 0x%04x\n", pfx,
321 pcie->command, pcie->status);
322 if (pcie->validation_bits & CPER_PCIE_VALID_DEVICE_ID) {
324 printk("%s""device_id: %04x:%02x:%02x.%x\n", pfx,
325 pcie->device_id.segment, pcie->device_id.bus,
326 pcie->device_id.device, pcie->device_id.function);
327 printk("%s""slot: %d\n", pfx,
328 pcie->device_id.slot >> CPER_PCIE_SLOT_SHIFT);
329 printk("%s""secondary_bus: 0x%02x\n", pfx,
330 pcie->device_id.secondary_bus);
331 printk("%s""vendor_id: 0x%04x, device_id: 0x%04x\n", pfx,
332 pcie->device_id.vendor_id, pcie->device_id.device_id);
333 p = pcie->device_id.class_code;
334 printk("%s""class_code: %02x%02x%02x\n", pfx, p[0], p[1], p[2]);
336 if (pcie->validation_bits & CPER_PCIE_VALID_SERIAL_NUMBER)
337 printk("%s""serial number: 0x%04x, 0x%04x\n", pfx,
338 pcie->serial_number.lower, pcie->serial_number.upper);
339 if (pcie->validation_bits & CPER_PCIE_VALID_BRIDGE_CONTROL_STATUS)
341 "%s""bridge: secondary_status: 0x%04x, control: 0x%04x\n",
342 pfx, pcie->bridge.secondary_status, pcie->bridge.control);
345 static void cper_estatus_print_section(
346 const char *pfx, const struct acpi_generic_data *gdata, int sec_no)
348 uuid_le *sec_type = (uuid_le *)gdata->section_type;
352 severity = gdata->error_severity;
353 printk("%s""Error %d, type: %s\n", pfx, sec_no,
354 cper_severity_str(severity));
355 if (gdata->validation_bits & CPER_SEC_VALID_FRU_ID)
356 printk("%s""fru_id: %pUl\n", pfx, (uuid_le *)gdata->fru_id);
357 if (gdata->validation_bits & CPER_SEC_VALID_FRU_TEXT)
358 printk("%s""fru_text: %.20s\n", pfx, gdata->fru_text);
360 snprintf(newpfx, sizeof(newpfx), "%s%s", pfx, INDENT_SP);
361 if (!uuid_le_cmp(*sec_type, CPER_SEC_PROC_GENERIC)) {
362 struct cper_sec_proc_generic *proc_err = (void *)(gdata + 1);
363 printk("%s""section_type: general processor error\n", newpfx);
364 if (gdata->error_data_length >= sizeof(*proc_err))
365 cper_print_proc_generic(newpfx, proc_err);
367 goto err_section_too_small;
368 } else if (!uuid_le_cmp(*sec_type, CPER_SEC_PLATFORM_MEM)) {
369 struct cper_sec_mem_err *mem_err = (void *)(gdata + 1);
370 printk("%s""section_type: memory error\n", newpfx);
371 if (gdata->error_data_length >= sizeof(*mem_err))
372 cper_print_mem(newpfx, mem_err);
374 goto err_section_too_small;
375 } else if (!uuid_le_cmp(*sec_type, CPER_SEC_PCIE)) {
376 struct cper_sec_pcie *pcie = (void *)(gdata + 1);
377 printk("%s""section_type: PCIe error\n", newpfx);
378 if (gdata->error_data_length >= sizeof(*pcie))
379 cper_print_pcie(newpfx, pcie, gdata);
381 goto err_section_too_small;
383 printk("%s""section type: unknown, %pUl\n", newpfx, sec_type);
387 err_section_too_small:
388 pr_err(FW_WARN "error section length is too small\n");
391 void cper_estatus_print(const char *pfx,
392 const struct acpi_generic_status *estatus)
394 struct acpi_generic_data *gdata;
395 unsigned int data_len, gedata_len;
400 severity = estatus->error_severity;
401 if (severity == CPER_SEV_CORRECTED)
402 printk("%s%s\n", pfx,
403 "It has been corrected by h/w "
404 "and requires no further action");
405 printk("%s""event severity: %s\n", pfx, cper_severity_str(severity));
406 data_len = estatus->data_length;
407 gdata = (struct acpi_generic_data *)(estatus + 1);
408 snprintf(newpfx, sizeof(newpfx), "%s%s", pfx, INDENT_SP);
409 while (data_len >= sizeof(*gdata)) {
410 gedata_len = gdata->error_data_length;
411 cper_estatus_print_section(newpfx, gdata, sec_no);
412 data_len -= gedata_len + sizeof(*gdata);
413 gdata = (void *)(gdata + 1) + gedata_len;
417 EXPORT_SYMBOL_GPL(cper_estatus_print);
419 int cper_estatus_check_header(const struct acpi_generic_status *estatus)
421 if (estatus->data_length &&
422 estatus->data_length < sizeof(struct acpi_generic_data))
424 if (estatus->raw_data_length &&
425 estatus->raw_data_offset < sizeof(*estatus) + estatus->data_length)
430 EXPORT_SYMBOL_GPL(cper_estatus_check_header);
432 int cper_estatus_check(const struct acpi_generic_status *estatus)
434 struct acpi_generic_data *gdata;
435 unsigned int data_len, gedata_len;
438 rc = cper_estatus_check_header(estatus);
441 data_len = estatus->data_length;
442 gdata = (struct acpi_generic_data *)(estatus + 1);
443 while (data_len >= sizeof(*gdata)) {
444 gedata_len = gdata->error_data_length;
445 if (gedata_len > data_len - sizeof(*gdata))
447 data_len -= gedata_len + sizeof(*gdata);
448 gdata = (void *)(gdata + 1) + gedata_len;
455 EXPORT_SYMBOL_GPL(cper_estatus_check);